Article: Prefab homes from Cover are designed by computer algorithms

06 Apr 2017

Specializing in backyard studios

If you’re in the market for a prefab dwelling—either as a full-time home or backyard unitoptions are aplenty. What L.A.-based startup Cover wants to add to the equation is a tech-driven efficiency that makes the whole design and building process a total breeze for the customer.

As detailed in a new profile on the company over on Co.Design, Cover sees itself as more of a tech company than a prefab builder. Indeed, whereas a typical prefab buying process would begin with choosing one of a few model plans and maybe then consulting with architects to tweak the design for specific needs, Cover turns the whole design process over to computer algorithms. Co.Design explains:

Once customers begin the design process, Cover sends them a survey of about 50 to 100 questions to inform the design. It asks about lifestyle–how many people typically cook a meal and what appliances are must-haves?–and structural needs, like should they optimize one view and block another one?

The company also use computer modeling to optimize window placement, cross-ventilation, and natural light, making use of zoning, sun-path, and geospatial data. All of these parameters are then sent to a proprietary computer program that spits out hundreds of designs that satisfy the requirements supplied.

Here are a couple of key things to know about Cover’s prefabs:

  • The company is specializing in the accessory dwelling unit, which is a secondary structure on a property with an existing single-family house. They can serve as guesthouses, in-law units, offices, yoga studios, and potentially a source of rental income.
  • While the computer will churn out a whole bunch of designs, Cover dwellings generally have a minimal modern look with an insulated steel structure, glass walls, and built-in storage.
  • When you order with Cover, the company takes care of the whole process, from coming up with a design, as described above (which takes three business days and $250), to acquiring necessary permits (two to five months, $20,000), to building and installation (12 weeks, final price contingent on the specific design). Some sample costs offered on the website are as follows: $70,000 for a guest room, $130,000 for a studio with a kitchenette, $160,000 for a one-bedroom unit, and $250,000 for a two-bedroom unit.

Via: Co.Design

Tags: #mscedc

April 06, 2017 at 11:40PM

 

Article: We Just Created an Artificial Synapse That Can Learn Autonomously

We Just Created an Artificial Synapse That Can Learn Autonomously

 

A team of researchers has developed artificial synapses that are capable of learning autonomously and can improve how fast artificial neural networks learn.

Mimicking the Brain

Developments and advances in artificial intelligence (AI) have been due in large part to technologies that mimic how the human brain works. In the world of information technology, such AI systems are called neural networks. These contain algorithms that can be trained, among other things, to imitate how the brain recognizes speech and images. However, running an Artificial Neural Network consumes a lot of time and energy.

Image Credit: Sören Boyn/CNRS/Thales physics joint research unit

Now, researchers from the National Center for Scientific Research (CNRS) in Thales, the University of Bordeaux in Paris-Sud, and Evry have developed an artificial synapse called a memristor directly on a chip. It paves the way for intelligent systems that required less time and energy to learn, and it can learn autonomously.

In the human brain, synapses work as connections between neurons. The connections are reinforced and learning is improved the more these synapses are are stimulated. The memristor works in a similar fashion. It’s made up of a thin ferroelectric layer (which can be spontaneously polarized) that is enclosed between two electrodes. Using voltage pulses, their resistance can be adjusted, like biological neurons. The synaptic connection will be strong when resistance is low, and vice-versa. The memristor’s capacity for learning is based on this adjustable resistance.

Better AI

AI systems have developed considerably in the past couple of years. Neural networks built with learning algorithms are now capable of performing tasks which synthetic systems previously could not do. For instance, intelligent systems can now compose music, play games and beat human players, or do your taxes. Some can even identify suicidal behavior, or differentiate between what is lawful and what isn’t.

This is all thanks to AI’s capacity to learn, the only limitation of which is the amount of time and effort it takes to consume the data that serve as its springboard. With the memristor, this learning process can be greatly improved. Work continues on the memristor, particularly on exploring ways to optimize its function. For starters, the researchers have successfully built a physical model to help predict how it functions. Their work is published in the journal Nature Communications.

Soon, we may have AI systems that can learn as well as out brains can — or even better

Author Dom Galeon April 5, 2017

Tags: #mscedc
April 06, 2017 at 03:24PM
Open in Evernote

Article: Unpaywall Is New Tool For Accessing Research Papers For Free

“Unpaywall” Is New Tool For Accessing Research Papers For Free

April 5, 2017 by Larry Ferlazzo

As anyone who has tried to pursue even a little bit of academic research can attest, publishers charge an arm-and-a-leg to access studies if you are not part of an institution that subscribes to their journals. And the authors of those studies don’t even get any of that money!

Last year, Sci-Hub broke through that barrier in one attempt (which may or may not be legal) to create more access – see The Best Commentaries On Sci-Hub, The Tool Providing Access to 50 Million Academic Papers For Free.

Today, another option was unveiled.

Today we’re launching a new tool to help people read research literature, instead of getting stuck behind paywalls. It’s an extension for Chrome and Firefox that links you to free full-text as you browse research articles. Hit a paywall? No problem: click the green tab and read it free!

The extension is called Unpaywall, and it’s powered by an open index of more than ten million legally-uploaded, open access resources.

Apparently, many institutions now require their faculty upload their published papers to their libraries, and that is a primary source for Unpaywall research.

I just tried it and it seems to work fairly well…

Tags: #mscedc
April 06, 2017 at 03:20PM

Impressive Adobe Algorithm Transfers One Photos Style Onto Another

Impressive Adobe Algorithm Transfers One Photo’s Style Onto Another


Mar 29, 2017

 

Two pairs of researchers from Cornell University and Adobe have teamed up and developed a “Deep Photo Style Transfer” algorithm that can automatically apply the style (read: color and lighting) of one photo to another. The early results are incredibly impressive and promising.

The software is an expansion on the tech used to transfer painting styles like Monet or Van Gogh to a photograph like the app Prisma. But instead of a painting, this program uses other photographs for reference.

“This paper introduces a deep-learning approach to photographic style transfer that handles a large variety of image content while faithfully transferring the reference style,” says the rather technical abstract of the Deep Photo Style Transfer paper.

Put more plainly: when you put in two photographs, the neural network-powered program analyzes the color and quality of light in the reference photo, and pastes that photo’s characteristics onto the second. This includes things like weather, season, and time of day—theoretically, a winter’s day can be turned into summer, or a cloudy day into a glorious sunrise.

The team’s early examples show the program in action. So this original photo:

Plus this reference photo:

Equals this final photo:

It’s important to note that the software does not alter the structure of the photo in any way, so there’s no risk of distorting the lines, edges or perspective. The entire focus is on mimicking the color and light in order to copy the “look” or “style” of a reference photograph onto a new shot.

Since this is a lot easier said than done, the program has to intelligently compensate for differences between the donor and receiving image. If there is less sky visible in the receiving image, it will detect this difference and not cause the sky to spill over into the rest of the original shot, for example.

The software even attempts to “achieve very local drastic effects,” such as turning on the lights on individual skyscraper windows, all without altering the original photo by moving windows around or distorting edges.

In the future, a perfected version of this technology could make its way into Photoshop as a tool, or run as a separate program or plug-in. Not that you should bank on this tech fixing the photos from your upcoming trip; like any other new technology, there is work to be done.

“The study shows that our algorithm produces the most faithful style transfer results more than 80% of the time,” the paper cautions. So maybe you can’t change Ansel Adam’s Moonrise, Hernandez to a Sunrise, Hernandez, but you get the picture (no pun intended) and it is very promising.

Tags: #mscedc

March 30, 2017 at 12:10AM

Article: Google’s new algorithm shrinks JPEG files by 35 percent

 

Google’s new algorithm shrinks JPEG files by
35 percent
David Lumb/17 Mar 2017

For obvious reasons, Google has a vested interest in reducing the time it takes to load websites and services. One method is reducing the file size of images on the internet, which they previously pulled off with the WebP format back in 2014, which shrunk photos by 10 percent. Their latest development in this vein is Guetzli, an open-source algorithm that encodes JPEGs that are 35 percent smaller than currently-produced images.

As Google points out in its blog post, this reduction method is similar to their Zopfli algorithm that shrinks PNG and gzip files without needing to create a new format. RNN-based image compression like WebP, on the other hand, requires both client and ecosystem to change to see gains at internet scale.

If you want to get technical, Guetzli (Swiss German for “cookie”) targets the quantization stage of image compression, wherein it trades visual quality for a smaller file size. Its particular psychovisual model (yes, that’s a thing) “approximates color perception and visual masking in a more thorough and detailed way than what is achievable” in current methods. The only tradeoff: Guetzli takes a little longer to run than compression options like libjpeg. Despite the increased time, Google’s post assures that human raters preferred the images churned out by Guetzli. Per the example below, the uncompressed image is on the left, libjpeg-shrunk in the center and Guetzli-treated on the right.

Tags: #mscedc
March 17, 2017 at 04:10PM